Path: utzoo!news-server.csri.toronto.edu!cs.utexas.edu!usc!bbn.com!drilex!dricejb
From: dricejb@drilex.UUCP (Craig Jackson drilex1)
Newsgroups: comp.arch
Subject: Re: Novice question:  measuring speed
Message-ID: <24780@drilex.UUCP>
Date: 16 Mar 91 22:36:53 GMT
References: <645@ssdc?> <3516:Mar1319:50:3291@kramden.acf.nyu.edu>
Organization: DRI/McGraw-Hill, Lexington, MA
Lines: 43

In article <3516:Mar1319:50:3291@kramden.acf.nyu.edu> brnstnd@kramden.acf.nyu.edu (Dan Bernstein) writes:
>In article <645@ssdc?> jbasara@ssdc (jim basara) writes:
>> Could someone please
>> provide me with a descriptive explaination of why MIP ratings are not a 
>> good way of comparing processing speed for RISC machines as opposed to MFLOPS?
>
>I've got this Turing machine that runs at 5000 MIPS. That's right, it
>can move the tape back and forth 5 *billion* times a second. Impressed?
>You shouldn't be: it takes so many instructions to get something done on
>a Turing machine that the MIPS measurement is pointless.

This should go in a FAQ list (Frequently Asked Questions) for comp.arch &
comp.benchmarks:

MIPS (Millions of Instructions Processed per Second) has nothing to do with
the number of instructions your machine processes in one second, unless your
machine is an IBM 370.  The bronzed 1-MIPS standard is a 370/158, I believe.

Over 10 years ago, Digital found that the VAX 11/780 had processing power
similar to that of the 370/158.  They used this information in their marketing
literature, and everyone began calling the 11/780 a 1-MIPS machine.  Actually,
it only processed about 500,000 instructions/second, on average.  However,
due to the nature of the instructions, it performed about the same work as
the 370/158 in a given unit of time, running some set of benchmarks.
(Don't ask me what they were; I know this equivalence predated Dhrystone.)

Today, when somebody says they have a 25-MIPS machine, it means "processes
some benchmark or benchmarks 25 times faster than an 11/780", unless
they are in the IBM world.

Now, as to why people quote MIPS instead of MFLOPS: Many uses to which a
computer can be put do not involve significant floating-point.  In particular,
in the business processing world where the MIPS measurement originated,
floating-point is unusual.  Therefore, people care more about the speed
of processing a non-floating-point instruction mix.

In the scientific world, floating-point is important.  The Usenet readership
tends to be more oriented towards floating-point computing than the computing
community at large.  You will find MFLOPS numbers quoted here from time to time.
-- 
Craig Jackson
dricejb@drilex.dri.mgh.com
{bbn,axiom,redsox,atexnet,ka3ovk}!drilex!{dricej,dricejb}